In 2013, Cody Wilson, a Texan law student with anarchist leanings, fired a pistol made almost entirely from 3D-printed plastic. The gun, called the Liberator, worked for precisely one shot. It also fired the starting gun on a decade of breathless predictions about additive manufacturing. Entire factories would soon become obsolete, we were told. Consumers would print everything from furniture to pharmaceuticals at home. (And even those home would be 3D-printed).
Wilson’s single-shot plastic pistol proved oddly prophetic. Not as a harbinger of decentralised weapon production, but as a metaphor for 3D-printing technology itself. Impressive in concept but limited in practice. A novelty rather than a transformation.
Global spending on 3D printing reached around 22 billion US dollars in 2024, according to Wohlers Associates, a consultancy that has tracked the industry since its earliest days. That sounds substantial until you set it against the 15 trillion US dollars global manufacturing sector. After three decades of development, additive manufacturing accounts for roughly 0.15 per cent of all the things that humans make. The machine shops and assembly lines that 3D printing enthusiasts pronounced doomed continue working.
Mass production keeps winning
Several problems have prevented 3D printing from scaling, with its lack of speed the most obvious. A traditional injection-moulding machine can spit out hundreds of identical parts per hour, while even the fastest commercial 3D printers manage just a handful.
For anything requiring volume, and most manufacturing requires volume, 3D printing simply isn’t viable. Desktop Metal, once among the most promising of the 3D printing start-ups, saw its market capitalisation collapse from billions of US dollars in early 2021 to an acquisition by Nano Dimension for 179 million US dollars in April 2025. Investors discovered that wanting a technology to succeed cannot make it economical.
Materials present another constraint. The polymers and metal powders used in additive manufacturing cost vastly more than conventional raw materials. A kilogram of 3D-printing resin runs to hundreds of dollars; injection-moulding pellets cost a few dollars at most. The machines themselves carry hefty price tags too. Industrial-grade metal printers start at around 500,000 US dollars and climb into the millions. Even with those investments, the finished products often require extensive post-processing such as sanding, curing, and machining that erodes any theoretical savings.
Then there is the question of what exactly people want to print. The home-manufacturing utopia imagined in the early 2010s assumed consumers would embrace making their own objects. They haven’t. The same impulse that drives people to assemble IKEA furniture rather than build tables from lumber extends to most household goods. Mass production won on efficiency a century ago and keeps winning.
The 3D niche
None of this, however, means that 3D printing is a technological dead end. The technology has carved out genuine niches where it makes sense. Prototyping is the most established. Design teams that once waited weeks for tooling can now iterate physical models in days. General Electric’s aviation division uses additive manufacturing to produce fuel nozzles for its LEAP jet engine, a part so geometrically complex that it previously required 20 separate components welded together. Printed as a single piece, it weighs 25 per cent less and runs five times longer. Aerospace tolerates the expense because weight savings translate directly to fuel savings over an aircraft’s operating life.
Medical applications show similar promise. Surgeons now routinely use 3D-printed models of patient anatomy to plan complex operations. Custom implants—titanium jaw reconstructions, cranial plates, hip sockets—can be fabricated to fit individual patients precisely. The dental industry has embraced the technology for crowns, bridges, and aligners. These are applications where customisation matters more than cost, and where traditional manufacturing would struggle regardless.
The construction industry, meanwhile, has moved from breathless announcements to quiet pragmatism. ICON, an Austin-based startup, has printed nearly 200 homes and structures across the US and Mexico, including the world’s largest 3D-printed community in Georgetown, Texas. A 372-square-metres two-storey home in Houston—the largest 3D-printed residential building in America—was completed using a COBOD printer. These structures cost less to build and waste less material than conventional methods. But they remain curiosities rather than commodities. The bottleneck is not the printing itself but everything else—site preparation, plumbing, electrical work, finishing. Even with a robot extruding concrete walls, most of the labour in construction remains stubbornly human.
Complementing, not replacing
This likely leaves the future of 3S printing technology as just one component in a more diverse manufacturing toolkit rather than its replacement. Hybrid approaches, such as printing complex geometries and machining critical surfaces, are gaining ground in aerospace and automotive applications. Materials science continues to advance. Carbon, a Silicon Valley firm, has developed a process called digital light synthesis that prints at speeds closer to conventional manufacturing. HP, a computing giant that entered the 3D printing market in 2016, continues to expand its Multi Jet Fusion technology with new materials and collaborations.
The comparison with personal computers is instructive. Early evangelists predicted PCs would eliminate paper, decentralise office work, and bring computing to every home. They were partly right, but the transformation took decades and looked nothing like the original prophecies. Manufacturing may follow a similar arc, slower and stranger than anyone predicted, but real.
3D printing is in no way a busted flush. It is a specialised tool that overheated marketing once compared to a new industrial revolution. That revolution has been postponed, perhaps indefinitely. The technology persists where geometry matters more than throughput, where customisation beats standardisation, and where the alternative is genuinely worse. The gun, so to speak, still works. It just fires less often than anyone hoped.
Photo: Dreamstime.
